Domain

Learning Health System

Type

Case Study or Comparative Case Study

Theme

notapplicable

Start Date

7-6-2014 1:15 PM

End Date

7-6-2014 2:45 PM

Structured Abstract

Lessons Learned from Recovery Act Comparative Effectiveness Research Data Infrastructure Investments

Thematic Domain: Learning Health Systems

Submission Type: Case Studies

Topic Area: NA

Introduction

The American Recovery and Reinvestment Act of 2009 directed approximately $300 million to data infrastructure (DI) projects that would develop new or enhance existing data resources for comparative effectiveness research (CER). Our study sought to describe these investments and to identify facilitators and barriers to implementation as well as lessons learned for future DI investments.

Methods

We reviewed redacted project proposals (97), conducted an investigator survey (70 investigators responded), and interviewed 11 project officers and more than 60 principal investigators. We classified DI investments into subgroups based on project characteristics and procurement source. We collected project characteristics in an Excel database to describe the investments and coded notes from interviews to identify facilitators, barriers, and lessons.

Findings

Over 70 percent of the DI projects assembled or enhanced existing clinical data sets and about 10 percent developed new data bases for CER. About two-thirds used data sets containing 100,000 or more covered lives, establishing linkages between various public and private data sources. About 58 percent of the projects built infrastructure or addressed research questions specific to one or more medical conditions, such as cardiovascular disease and cancer (13 and 9 projects, respectively). About half these investments proposed to bring together multiple data sources, and half also sought to leverage electronic health records (EHRs). The projects covered much geographical ground as well, with 40 percent using data sets covering multiple states and nearly a quarter using nationally representative data sets such as Medicare or Medicaid Analytic eXtract data. A large number (62) created data solutions and methodologies to bring together data beyond simply linking data sources.

Many projects built upon existing infrastructure with a foundation of informatics, network architecture, clinical research, and other fields of expertise. However, linking data from different systems, missing data, and obtaining data use agreements were more challenging than expected and caused project delays in some cases.

Lessons Learned

Many CER DI projects (about 80 percent) involved linking data held by different organizations, such as administrative claims data from health plans and EHR data from practices. Therefore, investigators found that building on existing relationships across organizations, such as clinical sites, hospitals and government agencies, helped them implement projects quickly. Investigators also noted that data collection was facilitated by making collection as seamless as possible for busy clinician practices. Researchers emphasized the development of DI that is readily understandable to clinicians, meets their needs, and fits into their clinical workflow.

Conclusions and Next Steps

Findings point to a number of potential recommendations for HHS as it considers future DI investments. First, to prioritize resources to build CER DI with the clinical detail needed to observe the effectiveness of different treatment strategies in diverse patient populations; second, to promote the development of clinical data resources that can provide CER data infrastructure but also serve other complementary purposes that enhance the value to providers; third, to support the ongoing costs of successful CER DI, such as updating databases and managing access to data.

Acknowledgements

Office of the Assistant Secretary for Planning and Evaluation

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Share

COinS
 
Jun 7th, 1:15 PM Jun 7th, 2:45 PM

Lessons Learned from Recovery Act Comparative Effectiveness Research Data Infrastructure Investments

Lessons Learned from Recovery Act Comparative Effectiveness Research Data Infrastructure Investments

Thematic Domain: Learning Health Systems

Submission Type: Case Studies

Topic Area: NA

Introduction

The American Recovery and Reinvestment Act of 2009 directed approximately $300 million to data infrastructure (DI) projects that would develop new or enhance existing data resources for comparative effectiveness research (CER). Our study sought to describe these investments and to identify facilitators and barriers to implementation as well as lessons learned for future DI investments.

Methods

We reviewed redacted project proposals (97), conducted an investigator survey (70 investigators responded), and interviewed 11 project officers and more than 60 principal investigators. We classified DI investments into subgroups based on project characteristics and procurement source. We collected project characteristics in an Excel database to describe the investments and coded notes from interviews to identify facilitators, barriers, and lessons.

Findings

Over 70 percent of the DI projects assembled or enhanced existing clinical data sets and about 10 percent developed new data bases for CER. About two-thirds used data sets containing 100,000 or more covered lives, establishing linkages between various public and private data sources. About 58 percent of the projects built infrastructure or addressed research questions specific to one or more medical conditions, such as cardiovascular disease and cancer (13 and 9 projects, respectively). About half these investments proposed to bring together multiple data sources, and half also sought to leverage electronic health records (EHRs). The projects covered much geographical ground as well, with 40 percent using data sets covering multiple states and nearly a quarter using nationally representative data sets such as Medicare or Medicaid Analytic eXtract data. A large number (62) created data solutions and methodologies to bring together data beyond simply linking data sources.

Many projects built upon existing infrastructure with a foundation of informatics, network architecture, clinical research, and other fields of expertise. However, linking data from different systems, missing data, and obtaining data use agreements were more challenging than expected and caused project delays in some cases.

Lessons Learned

Many CER DI projects (about 80 percent) involved linking data held by different organizations, such as administrative claims data from health plans and EHR data from practices. Therefore, investigators found that building on existing relationships across organizations, such as clinical sites, hospitals and government agencies, helped them implement projects quickly. Investigators also noted that data collection was facilitated by making collection as seamless as possible for busy clinician practices. Researchers emphasized the development of DI that is readily understandable to clinicians, meets their needs, and fits into their clinical workflow.

Conclusions and Next Steps

Findings point to a number of potential recommendations for HHS as it considers future DI investments. First, to prioritize resources to build CER DI with the clinical detail needed to observe the effectiveness of different treatment strategies in diverse patient populations; second, to promote the development of clinical data resources that can provide CER data infrastructure but also serve other complementary purposes that enhance the value to providers; third, to support the ongoing costs of successful CER DI, such as updating databases and managing access to data.